Practice Uncertainties

What is the most effective treatment for frozen shoulder?

BMJ 2016; 354 doi: https://doi.org/10.1136/bmj.i4162 (Published 23 August 2016) Cite this as: BMJ 2016;354:i4162
  1. Amar Rangan, professor of orthopaedic surgery1 2 3,
  2. Nigel Hanchard, reader in orthopaedics4,
  3. Catriona McDaid, senior research fellow3
  1. 1The James Cook University Hospital, Middlesbrough TS4 3BW, UK
  2. 2Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford OX3 7HE, UK
  3. 3York Trials Unit, Department of Health Sciences, University of York, York YO10 5DD, UK
  4. 4School of Health and Social Care, Teesside University, Middlesbrough TS1 3BX, UK
  1. Correspondence to: A Rangan amar.rangan{at}york.ac.uk

What you need to know

  • Consider the diagnosis in patients with pain and limitation of passive external rotation of the shoulder

  • There is insufficient evidence to reliably recommend a treatment approach. Moderate evidence supports corticosteroid injection or hydrodilatation; and physiotherapy in reduction of pain and stiffness

  • Consider specialist referral if the patient does not respond to conservative treatment or if the diagnosis is in doubt

Frozen shoulder is a painful condition that most commonly affects people in their 50s,1 2 and may limit daily activities and disturb sleep.3 Its annual prevalence has been estimated as 1.4 per 1000 patients in a UK study of two general practices (total population 17 000).4

Arthroscopic view of frozen shoulder, showing the humeral head (black arrowhead), glenoid (white arrow), and rotator interval with inflamed and proliferative tissue (black arrow) and site for injection

The diagnosis is clinical (see box 1). Restriction or pain, or both, on shoulder elevation and external rotation have been consistently described in the literature,6 8 9 11 12 yet there are insufficient data to evaluate diagnostic accuracy of signs and symptoms.6 Restriction and pain are evident on passive as well as active testing, which differentiates it from rotator cuff disease (where passive movements are classically of full range).6 Another differentiator is crepitus in the shoulder, which is more indicative of arthritis.12 Three phases (pain, stiffness, and resolution) are described, but they often overlap. Patients may have “pain-predominant” or “stiffness-predominant” frozen shoulder.13

Box 1: Diagnostic criteria for frozen shoulder567891011

The diagnosis is clinical:

  • Pain arising insidiously in deltoid region with increasing shoulder stiffness

  • Pain at the end of external rotation, and restriction of this movement on both active and passive testing

  • No obvious crepitus on movement

X rays are not routinely required, but if obtained, should show a normal glenohumeral joint

Although frozen shoulder is considered …

View Full Text

Sign in

Log in through your institution

Free trial

Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
Sign up for a free trial

Subscribe